electrum commented on pull request #1508: URL: https://github.com/apache/iceberg/pull/1508#issuecomment-700794725
The question is what to do when we access a historical snapshot for older snapshots. It seems we have a few choices: 1. Use the current schema, as we do today. 2. Fail the query. 3. Fallback to reading old metadata files. Using the current schema seems like the worst choice, as that results in inconsistent behavior for reasons that are invisible to the user. Failing the query would break behavior that works today. While I agree that having more code to handle older data is annoying, the complexity here seems low (~20 lines of code) relative to the benefit. We have much higher complexity in other areas, such as supporting non-projected identity columns for converted Hive tables (and that's not even covered by the specification). ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
